Respiratory strategy at birth initiates distinct lung injury phenotypes in the preterm lamb lung.

Background: A lack of clear trial evidence often hampers clinical decision-making during support of the preterm lung at birth. Protein biomarkers have been used to define acute lung injury phenotypes and improve patient selection for specific interventions in adult respiratory distress syndrome. The objective of the study was to use proteomics to provide a deeper biological understanding of acute lung injury phenotypes resulting from different aeration strategies at birth in the preterm lung.

Methods: Changes in protein abundance against an unventilated group (n = 7) were identified via mass spectrometry in a biobank of gravity dependent and non-dependent lung tissue from preterm lambs managed with either a Sustained Inflation (SI, n = 20), Dynamic PEEP (DynPEEP, n = 19) or static PEEP (StatPEEP, n = 11). Ventilation strategy-specific pathways and functions were identified (PANTHER and WebGestalt Tool) and phenotypes defined using integrated analysis of proteome, physiological and clinical datasets (MixOmics package).

Results: 2372 proteins were identified. More altered proteins were identified in the non-dependent lung, and in SI group than StatPEEP and DynPEEP. Different inflammation, immune system, apoptosis and cytokine pathway enrichment were identified for each strategy and lung region. Specific integration maps of clinical and physiological outcomes to specific proteins could be generated for each strategy.

Conclusions: Proteomics mapped the molecular events initiating acute lung injury and identified detailed strategy-specific phenotypes. This study demonstrates the potential to characterise preterm lung injury by the direct aetiology and response to lung injury; the first step towards true precision medicine in neonatology.